Structural topology optimization considering both manufacturability and manufacturing uncertainties
Chao Wang,Bin Xu,Zunyi Duan,Jianhua Rong
DOI: https://doi.org/10.1007/s00158-022-03458-9
IF: 4.279
2022-12-29
Structural and Multidisciplinary Optimization
Abstract:This work proposes a robust and efficient approach to structural topology optimization considering both manufacturable connectivity and manufacturing uncertainties. It can be seen as an extension of the three-field robust method for compliance minimization based on eroded, intermediate, and dilated projections. The novelty of this proposal comes from the rational inclusion of the Poisson equation-based potential constraint for manufacturable connectivity in the three projected fields. This helps to achieve manufacturable designs with reliable performance and favorable machinability against uniform manufacturing errors. Notably, a meaningful potential law of the three projection-based connectivity designs is revealed. Accordingly, an effective potential constraint strategy is developed to reduce the heavy computational cost associated with the complicated robust optimization involving multiple design fields and nonlinear constraints. Also, an applicable solving scheme is provided to cope with numerical difficulties arising in the solution. Based on typical 2D and 3D examples, four optimization formulations related to the research topic are thoroughly compared to evaluate the performance of the proposed strategies. The results show that the design with purely manufacturable constraint is not immune to manufacturing errors. And, the combination of manufacturability and manufacturing uncertainties is feasible and necessary for manufacturable designs with reliable performance requirements.
mechanics,computer science, interdisciplinary applications,engineering, multidisciplinary